Leading memory theories emphasize that new learning occurs on the background of existing knowledge. Retrieving prior knowledge during new experiences allows new information to be integrated into existing memories, resulting in the formation of rich, cohesive memory networks that relate discrete events. This integration process is proposed to facilitate new learning and enable memories to extend beyond direct experience to anticipate the relationships among events. However, memory for evidence neurobiological inability to directly measure the contents of reactivated memories during new experiences. To address this critical gap, the proposed studies employ a new experimental paradigm that uses highly sensitive pattern classifier algorithms applied to functional neuroimaging data to quantify incidental memory reactivation during new event encoding. Quantifying memory reactivation allows us to test mechanistic predictions about how past memories influence learning in the present. Aim 1 will use this paradigm to test the hypothesis that hippocampus and ventromedial prefrontal cortex (VMPFC) work in concert to support memory integration during new learning. We propose that by linking new information with well-established memories, this hippocampal-VMPFC mediated encoding process improves new learning and enables novel judgments about relationships among distinct events. Aim 2 will examine how temporal context and memory strength influence the formation of integrated memory traces for related events. We propose that learning overlapping events within the same temporal context facilitates memory integration by enhancing memory reactivation and recruiting hippocampal-VMPFC encoding processes. We will also adjudicate between opposing theoretical perspectives of learning that make competing predictions for whether strong or weak memories lead to enhanced memory integration. Aim 3 will use high- resolution fMRI focused on the medial temporal lobe to determine the precise hippocampal computations and coding strategies that underlie memory integration. We will determine the relationship between memory reactivation and hippocampal mismatch responses that signal differences between current events and existing memories to test the hypothesis that mismatch responses trigger memory integration. We will also use pattern- information analysis to test the hypothesis that the hippocampus creates integrated memories by forming overlapping neural codes for related events. Collectively, this work will determine how internally generated content influences new learning and will isolate the precise neural networks, computations, and coding strategies that underlie memory integration. Understanding how the brain uses prior experience to make sense of new information will lay the foundation for translational work onintegration and its functional significance is sorely lacking due primarily to an effective for interventions therapeutic psychiatric and neurological disorders that require acquisition and maintenance of new behaviors.